Test Stimulus Optimization of Analog Circuits based on Deviation Synthesis

The paper aims at the problem of low incipient fault detection accuracy in analog circuits, and a method based on deviation synthesis is proposed to solve this problem. In this method, the characteristic deviation and slope deviation of the frequency response curves of normal state and fault state are synthesized at each test frequency point, and the synthetic deviation is taken as the evaluation index to select the test frequencies for the analog circuit. Furthermore, the optimal test frequency set is obtained accurately by using the Greedy algorithm to remove the redundant frequencies. The optimal test stimuli not only can improve the incipient fault detection accuracy of the analog circuit, but also can reduce the operation time for test frequency generation. Meanwhile, the multifrequency test stimuli set contains fewer elements. The simulation experiment results show that the optimal test stimuli can obtain higher incipient fault detection accuracy of the analog circuit, compared with the traditional pulse stimulus.